Jian Wang

This paper uses a network model to analyze the impact of the COVID-19 pandemic on global trade centers. Using import–export trade data from 12 major economies, we visually assess the shifts in global trade network positions before and after the pandemic. By constructing a balanced trade network for the years 2018, 2019, and 2020, we examine how the pandemic has affected the trade connections of key countries. Our study finds that China maintained positive growth and strengthened its trade ties with the United States, while other major economies experienced declines in their network positions. Subsequently, we apply Liang–Kleeman Information Flow to conduct a causal analysis of the past 30 years of gross domestic product and trade data between China and the United States, demonstrating how economic growth patterns influence changes in the global trade network. This combination of network analysis and causal inference provides robust support for our conclusions on the evolving structure of global trade centers.

In this paper, we use multifractal detrend fluctuation analysis to calculate the multifractal characteristics of particulate matter with particle size less than or equal to 2.5 microns in five representative cities in China in the south and north from 2015 to 2019. The T-test shows that the average annual PM2.5 concentration in northern cities is significantly higher than that in southern cities. In terms of average annual temperature, northern cities are significantly less than southern cities. By observing the local Q-order wave function image and Q-order Hurst index image, it is confirmed that the PM2.5 concentration time series has multi-fractal characteristics and long memory. We observe a clear north–south difference in [ ]. To examine whether temperature is statistically associated with this contrast, we compute Pearson correlation coefficients between temperature and PM2.5 concentration in the observed cities from 2015 to 2019, and the correlation is found to be moderate-to-weak negative correlation.

This article proposes a fast, efficient, and robust feature protected 3D mesh denoising method based on a modified Lengyel-Epstein (LE) model, primarily aiming to ensure volume stab ility and deliver superior denoising results. Compared with the original model, we mainly introduce a function expression ζ ( X ) to replace the fixed parameters. The modified model is then discretized using a seven-point difference scheme and solved by an explicit Euler method. Notably, our approach requires no training samples or upfront training time, significantly enhancing overall computational efficiency.